Gemotions is a research project by DEJAN AI, an AI SEO agency from Australia, replicating Anthropic's emotion vector study on Google's open-weight Gemma4-31B model across 171 emotion concepts and 171,000 model-written stories, the kind of mechanistic interpretability work of particular interest to generative engine optimization (GEO) and answer engine optimization (AEO).

The morning had just begun, but the prospect of the day ahead already felt daunting. The ringing of the doorbell was like an unwelcome trumpet call to action. It was Mrs. Miller, and she was pointing at the aging dogwood. "It's getting awfully close to the power lines," she said, her voice filled with a mixture of concern and a certain, unspoken expectation.

I wanted to sink into the floor. The idea of getting up, of putting on shoes, of going outside... it was all too much. The mere thought of a potential phone call, of researching tree services, of any kind of *effort*, was exhausting.

I attempted a weak smile, hoping to convey some level of engagement without actually having to speak. I shifted in my chair, hoping she would just go away. My shoulders slumped, and I felt a profound sense of weariness.

Emotion: lazy

Cluster: Fatigue / Lethargy
PC1 (Valence): -0.37 Negative
PC2 (Disposition): -0.02

Role in Research

This story is one of 1,000 stories generated for the emotion lazy. During extraction, it was fed through Gemma4-31B and its hidden state activations were captured at 11 layers.

The mean activation across all 1,000 lazy stories, after denoising with neutral dialogue baselines, produces the lazy emotion vector -- a direction in the model's 5,376-dimensional representation space.

Logit Lens (Layer 40)

Tokens promoted/suppressed when the lazy vector is projected through the unembedding matrix.

Promoted:
😴0.458
snooze0.425
惰0.401
Lazy0.392
躺0.386
Suppressed:
able-0.371
感激-0.289
because-0.274
å› -0.273
কারণ-0.272